dewar_note on the liquefaction of h2 and he

8/8/2019 Dewar_note on the Liquefaction of h2 and He

1/8

588 D E W A R : N O T E ON T H E LI QU EF A C TI ON O F

Ey Prof . JAJIICSEWAR.IN paper entitled The Liquefaction of Air and Research itt LowTemperatures, read before the Chemical Society and published in theProceedings, No. 158, an account is given of the history of the liquidhydrogen problem and th e result of my own experiments np to theend of the year 1835.

Wroblewski made the first conclusive experiments on the liquefac-tion of hydrogen in January, 1884. H e found t ha t the gas cooled ina capillary glass tube to the boiling point of oxygen, and expandedquickly from 100 to 1 atmos., showed t h e same appearance of suddenebullition, lasting for rz fraction of a second, as Cailletet had seenin his early oxygen experiments. No sooner had the announcementbeen made, than Olszewski confirmed th e result by expanding hydrogenfrom 190 atmos., previously cooled to the temperature given by liquidoxygen and nitrogen evaporating under diminished pressure. Olszewski,however, declared in 1884 th at he saw colourless drops, and by partialexpansion to 40 atmos. the liquid hydrogen was seen by him runningdown the tube. Wroblewski could not confirm Olszewskis results,his hydrogen being always obtained in the form of what he called a iquide dynamique, or the appearance of an instantaneous froth.The following extract from Wroblewskis paper (Conzpt. rend., 1885,100, 981) states very clearly the results of his work on hydrogen :-Lhydroghne soumis B la pression de 180 atm. jusquk 190 ntm.refroidi par lazote bouillant dans In vide (A la tempkrature de S Rsolidification) e t dktendu brusquement sous la pression atmospheriquepresente une mousse bien visible. De la couleur grise de cet te mousse,oh lceil ne peut distinguer des goutt elett es incolores, on ne peut pasencore deviner quelle apparence aurait lhydroghe a 18tat de liquidesta tique e t lon es t encore moins autorisk A prh ise r sil a o u non uneapparence mdtallique.

J a i pu placer dans cette mousse ma pile thermo-t5lectriqne e tjobtenu suivant les pressions employkes des tempkratures de -208jusqua -2 11 C. J e ne peux pas encore dire dans quelle relation setrouvent ces nombres avec la tempbrature rCele de la moussee t la t e m p h t u r e dkbullition de lhydrogkne soils la pressionatmosphdrique, puisque je nai pas encore la certitude que lafaible dur4e de ce phknomhne a i t permis h la pile de se refroidircomplhtement. Nbanmoins, je crois anjourdhui de mon devoir depnblier ces r&sultats, afin de prCcisei. ldtat actuel de la question

The facts are substantially as follows.


2/8

H Y D R O G E N A N D H E L I U M . 529de l a liquefaction de lhydroghne. It is well to note thatthe lowest thermo-electric temperature recorded by Wroblewskiduring the adiabatic expausion of thc hydrogen, namely, - 2 1 1 isreally equivalent to a much lower temperature on th e gas thermometerscale. The most probable value is - 230, and this must be re-garded as the bighest temperature of the liquid state, or the criticalpoint of hydrogen according to his observations. The above methodshaving failed to produce static hydrogen, Wroblewski suggested tha tthe result might be attained by the use of hydrogen gas as a coolingagent. From this time until his death in the year 1888, Wroblewskidevoted his time to a laborious research on the isothermals of hydrogena t low temperatures. The data thus arrived at enabled him, by theuse of Van der Waals formulae, to calculate the critical constants andboliing point of liquid hydrogen.

Olszewski returned t o the subject in 1891, repeating and correctinghis old experiments of 1884, which Wrobleski had failed to confirm, ina glass tube 7 m m . in diameter instead of one of 2 mm., as i n the earlytrials. H e says, On repeating my former experiments, I had no hopeof obtaining a lower temperature by means of any cooling agent, bu t Ihoped that the expansion of hydrogen would be more efficacious, onaccount of the larger scale on which the experiment was made. Theresult of these experiments Olszewski describes as follows, Thephenomenon of hydrogen ebullition, which was then observed, wasmuch more marked and much longer th an during my former investiga-tions in the 6ame direction. But even then I could not perceive a n ymeniscus of l iquid hydrogen. Further, The reason for which it hasnot been hi ther to possible to liquefy hydrogen i n a static state, is thatthere exists no gas having a density between those of hydrogen andof nitrogen, and which might be f o r instance 7-10 (H 1). Such a gascould be liquefied by means of liquid oxygen or air as cooling agent,and be afte rwards used as a frigorific menstruum in t h e liquefaction ofhydrogen.

Professor Olszewski, in 1895,determined the temperature reached inthe momentary adiabatic expansion of hydrogen a t low temperatures ;just as Wroblewski had done in 1885, only be employed a platinumresistance thermometer instead of a thermo-junction.

For thi s purpose, he used a small steel bott le of 20 or 30 C.C. capacity,containing a platinum resistance thermometer ; n th is way, tempera-tures were registered which were regarded as those of the critical andboiling points of liquid hydrogen, a substance which could n o t be seenunder the circumstances, and was only assumed a t the most to exist forx second o r two dur ing the expansion of the gaseous hydrogen in thesmall steel bottle.


3/8

530 DE W AR : NOTE .ON THE LIQUEFACTION OFThe results arrived at by Wroblewski and Olszewski are given in

the following table.Wroblewski, 1885. Olszewski, 1895.

Critical temperature. , . - 240 - 234- 243Ooiling point., ,...... .. - 250Critical pressure . ,. 13At. 20At.

The moment the critical point is approximately defined, the boilingpoint is roughly known, and the critical pressure can be inferred fromanalogy with the behaviour of other substances.lh a paper published in the Phil. Mag., September, 1884, On the

Liquefaction of Oxygen and the Critical Volumes of Fluids, thesuggestion was made th at th e critical pressure of hydrogen was wrong,and that instead of being 99 atmos. (as deduced by Sarrau fromAmagats isothermals), th e gas had probably an abnormally low valuefor this constant. This view was substantially confirmed by Wroblewskifinding a critical pressure of 13.3 atoms., or about one-fourth of thatof oxygen. The Chemical News (September 7, 1894) contains anaccount of the stage th e authors hydrogen experiments had reachedat th at date. The object was t o collect liquid hydrogen a t its boilingpoint, in an open vacuum vessel, which is a much more difficult problemthan seeing the liquid in a glass tube under pressure and at a highertemperature. I n order to raise the critical point of hydrogen toabout - 200 from 2 to 5 per cent. of nitrogen or air was mixed with it.This is simply making a n artificial gas containing a large proportionof hydrogen which is capable of liquefaction by the use of liquid air.The results are summed up in th e following extract from the paper.One thing can, however, be proved by the use of the gaseous mixtureof hydrogen and nitrogen, namely, th at by subjecting it to a high com-pression a t a temperature of - 200 and expanding the resulting liquidinto air, a much lower temperature than anyth ing th at has beenrecorded up to the present time can be reached. This is proved bythe fact tha t such a mixed gas gives, under the conditions, a paste orjelly of solid nitrogen, evidently giving off hydrogen, because the gascoming off burns fiercely. Even when hydrogen containing only some2 to 5 per cent. of air is similarly trea ted, the result is a white, solidmatte r (solid ai r) along with a clear liquid of low density, which isso exceedingly volatile that no known device for collecting has beensuccessful.

The eport of a Friday Evening Lecture on New Researches onLiquid Air (Proc . Roy. Inst., 1896) contains a drawing of the appa-ra tus employed for the production of a jet of hydrogen containingvisible liquid. A represents one cfthe hydrogen cylinders ; B and C, vacuum vessels containing carbonicacid under exhaustion and liquid air respectively; D is the coil, G the

This is reproduced in the figure.


4/8

HYDROUEN AND HELIUM. 531pin-hole nozzle, and F the valve. By means of thi s jet, liquid air canbe quickly transformed into a hard solid. It was shown that such ajet could be used to cool bodies below the temperature that it ispossible to reach by the use of liquid air, but all attempts tocollect the liquid hydrogen from the jet in vacuum vessels failed.No other investigator has, so far, improved on the results the authordescribed in the Proceedings of t he Chemical Society (No. 158),1895, or, indeed, touched the subject since tha t date. The type of

B

APPARATUSSED IN THE PRODUCT ION OF T H E LIQUIDH Y D R O G E NE T .apparatus used in these experiments worked well, so it was re-solved t o construct a much larger liquid ai r plant, and t o combinewith it circuits and arrangements for the liquefaction of hydrogen,which will be described in a subsequent paper. This apparatus tooka year to build up, and many months have been occupied in testingand making preliminary trials. The many failures and defeats neednot be detailed.

On May 10th of this year, starting with hydrogen cooled t o - 206",and under a pressure of 180 atmospheres, escaping continuously from


5/8

532 D E W A R : NOTE ON T H E LI QU EF A C TI ON OFthe nozzle of a coil of pipe a t the ra te of about 1 0 t o 15 cub. ft.per minute, in a vacuum vessel doubly silvered and of special con-struction, all surrounded with a space kept below -200, liquidhydrogen commenced to drop from th is vacuum vessel i nto anotherdoubly isolated by being surrounded with a third vacuum vessel.I n about 5 minutes, 20 cubic centimetres of liquid hydrogen werecollected, when the hydrogen jet froze up, from th e accumulation ofair in the pipes frozen out from the impure hydrogen, The yield ofliquid was about 1 per cent. of the gas. The hydrogen in the liquidcondition is clear and colourless, showing no absorption spectrum, andthe meniscus is a s well defined as in the case of liquid air. The liquidmust have a relatively high refractive index and dispersion, and thedensity appears also to be in excess of the theoretical density, namely,0.18 to 0.12, which we deduce respectively from the atomic volume oforganic compounds, and the limiting density found by Amagat forhydrogen gas under infinite compression. Ye t this may be a delusiondue t o its high dispersion. A preliminary at tem pt t o weigh a smallglass bulb in the liquid made the density about 0.08. M y oldexperiments on the density of hydrogen in palladium gave a value forthe combined element of 0 . 6 2 , and i t mill be interesting t o find theaccurate density of tho liquid substance at it s boiling point. Nothaving arrangements a t hand to determine the boiling point, otherthan a thermo-junction which gave entirely fallacious results, experi-ments were made to prove the excessively low temperature of theboiling fluid. I n the first place, if a long piece of glass tubing, sealedat one end and open to the air a t the other, is cooled by immersingthe closed end in the liquid hydrogen, the tube immediately fills,where it is cooled, with solid a i r ; a small tube containing liquidoxygen became a bluish solid. A first trial of putting the liquidhydrogen under exhaustion gave n o appearance of transition in to th esolid state. The liquid hydrogen in its vacuum tube, which is immersedin liquid ai r so th at th e external wall of the vacuum vessel is maintainedat about - 190 is found t o evaporate a t a ra te not fa r removed fromthat of liquid air from a similar vacuum vessel under the ordinaryconditions of storage. This leads me to the conclusion that , withproper isolation, it will be possible to manipulate with liquid hydrogenas easily as with liquid air . The second experiment was made with atube cortaining helium.

The Cracow Academy Bulletin for 1896 containsa paper by ProfessorOlszewski, enti tled, A Research on the Liquefaction of Helium, i nwhich he states, As far as my experiments go, helium remains a per-manent gas and apparently is much more difficult to liquefy thanhydrogen. In a paper of m y own in the Proceedings of the Chemical


6/8

HYDROGEN AND HELIUM. 533Society, No. 183 (1896-7), in which the separation of helium fromBath gas was effected by a liquefaction method, th e suggestion wasmade that the volatility of hydrogen and helium would probably befound close together, just like those of fluorine and oxygen. Havinga specimen of purified helium, which had been extracted from Bath gas,sealed up in a bulb with a narrow tube attached, the latte r was placedin liquid hydrogen, when a distinct liquid was seen to condense. Thesame experiment repeated, only using liquid air evaporating in nvacuum gave no trace of condensation. From this result, it wouldappear that there cannot be any great difference between the boilingpoints of helium and hydrogen.

All known gases have now been condensed in to liquids which canbe manipulated at their boiling points under atmospheric pressure insuitably arranged vacuum vessels. Wit h hydrogen as a cooling agentl,we shall get within 20 or 30 of the zero of absolute temperature andits use will open up an entirely new field of scientific inquiry. Evenas great a man as James Clerk Maxwell had doubts as to the possibilityof ever liquefying hydrogen, see Scienti jc Papem, 2, 412. I n con-cluding his lectures on the non-metallic elements, delivered at theRoyal Institution in 1852 and published the following year,Faraday said(See Faradays Lectures on the Non-metallic Elements p. 292-3),There is reason to believe we should derive much information as tothe intimate nature of these non-metallic elements, if me could succeedin obtaining hydrogen and nitrogen in the liquid or solid form. Manygases have been liquefied; th e carbonic acid gas has been solidified,bu t hydrogen and nitrogen have resisted all our efforts of the kind.Hydrogen in many of its relations acts as though it were a metal;could it be obtained in a liquid or solid condition, the doubt might besettled. This great problem, however, has yet to be solved ; or shouldwe look with hopelessness on th is solution when we reflect with wonder-and as I do almost with fear and trembling-on the powers of in-vestigating the hidden qualities of these elements-of questioningthem, making them disclose their secrets and tell their tales-givenby the Alm.ighty t o man.

Faradays expressed fa ith i n th e potentialities of experimental in-quiry in 1852 has been justified forty-six years afterwards by theproduction of liquid hydrogen in the very laboratory i n which all hisepoch-making researches were executed. The ( oubt has now beensettled, hydrogen does not possess in the liquid state the characteristicsof a metal. No one can predict th e properties of matte r near thezero of temperature. Faraday liquefied chlorine in the year 1823.Sixty years afterwards, Wroblewski and Olszewski produced liquid air,and now, after a fifteen years interval, the remaining gases, hydrogenand helium, appear as static liquids. Cocsidering the step from the

VOL. LXXIII. N N


7/8

534 D E W A R : NOTE ON T H E LIQUEFACTION O Fliquefaction of ai r to tha t of hydrogen is relatively as great in thethermodynamic sense as that from liquid chlorine to liquid air, thefact that the former result has been achieved in one-fourth the timeneeded to accomplish the latte r proves the greatly accelerated race ofscientific progress in our time,The efficient cultivation of this field of research depends upon com-bination and assistance of an exceptional kind, bu t in the first instancemoney must be available, and the members of the Royal Inst itut iondeserve my especial gratitude for their handsome donations to theconduct O F thi s research. Unfortunately, its prosecution will demanda fur ther large expenditure. It is my duty also to acknowledge that atan early stage of the enquiry the Hon. Company of Goldsmiths helpedlow temperatures investigation by a generous donation to the ResearchFund.During the whole course of the low temperature work, carried outa t the Royal Institution, the invaluable aid of Mr. Robert Lennoxhas been at my disposal, and it is not too much to say that , but forhis engineering skill, manipulative ability, and loyal perseverance, thepresent successful issue might have been indefinitely delayed. M ythanks are also due to Mr. J. W. Heath for valuable assistance in theconduct of these experiments,

Addendum.Since the above paper was written, both the boiling point andlspecific gravity of hydrogen have been determined. The boiling pointin the meantime given by the use of .a platinum resistance ther-mometer involves, however, extrapolation of the curve correlatingtemperature and resistance; the result is that the boiling point

of hydrogen is -238' C. or 35' absolute. At this temperature,the tension of liquid air (which, of course, becomes solid) is lessthan 0.002 mm. The resistance of the thermometer used was 5.338ohm at the melting point of ice, and this was reduced to 0.129ohms when placed in boiling hydrogen. The absolute zero inplatinum degrees of this thermometer was - 263.27, and the tem-perature measured on this scale is -256.29' or 6 0 3 8 ~rom thepoint where the conductivity of the platinum would become infinite.The resistance of the platinum in the liquid hydrogen is reduced tonearly ?=th of what it is in liquid oxygen. It will be necessary to findout the electric conductivity of the fluid itself, and to repeat th e obser-vations with other thermometers, before we can arrive at more definiteconclusions. The vapour of hydrogen at its boiling point is abouteight times denser t han the gas at ordinary temperatures, or it hasabout half the density of air, whilst the vapour coming off from liquid


8/8

HYDROGEN A N D HELIUM. 535air at its boiling point is somewhat less than f o u r times the density ofair a t the ordinary temperature. By evaporation in a vacuum, thetemperature of liquid hydrogen will be lowered from 1Q" o 1 5 O , but i twill be practically impossible (so far as we can anticipate the resultsof experiment) t o reach a lower temperature than - 250' c! or 20'absolute by this means. A t present, we can see no way of bridgingover the last 20 to 25 degrees, and therefore the approach to thezero of absolute temperature and the study of matter and energy undersuch conditions must be confined to temperatures above 25' absolute.

The density of liquid hydrogen has been approximately determined byevaporating some 10 C.C. of the liquid, and collecting and measuringthe gas produced, thereby ascertaining its weight. I n this way, 8.15litres at 1 4 O C. and 753 mm. were collected over water from between9 and 10 C.C. of liquid hydrogen. It appears, therefore, that the densityof the liquid is about 0.07,using whole nnmbers as the calculationworks out to 0.068 nearly. Liquid hydrogen is, therefore, a verydeceptive fluid so far as appearance goes. The fact of it s collecting soeasily, dropping so well, and having such a well defined meniscus inducedme to believe tha t the density might be about half that of liquidair. It was a great surprise to find the density only &th of water.Liquid marsn gas was the lightest known liquid, the density a t i ts boilingpoint being 0.417, but liquid hydrogen has only +th the density of thissubstance. The density of occluded hydrogen in palladium being 0.62,i t is eight times denser than the liquid.

Hydrogen in the liquid state is 100 times denser than the vapouri t is giving off a t its boiling point, whereas liquid oxygen is 255 timesdenser than its vapour. It appears, therefore, that the atomic volumeof liquid hydrogen a t its boiling point is 14.3, as compared with 13.7for oxygen under similar circumstances. In other words, they arenearly identical. From this we can infer that the critical pressure neednot exceed 15 atmospheres. The extraordinary properties theoryrequires hydrogen ahould possess, especially as regards specific andlatent heat, become more intelligible from the moment we know that thedensity is:so small. I n other words, when we compare the propertiesof equal volumes of liquid hydrogen and ai r under similar correspondingtemperatures, they do not differ more than might be anticipated.

ROYAL NSTITUTION.

dewar_note on the liquefaction of h2 and he

Documents